CN107085850B - The method and book scanning method that masking acquires the marked body of foreign matter in image, identifies foreign matter marked body in image - Google Patents
The method and book scanning method that masking acquires the marked body of foreign matter in image, identifies foreign matter marked body in image Download PDFInfo
- Publication number
- CN107085850B CN107085850B CN201710182739.2A CN201710182739A CN107085850B CN 107085850 B CN107085850 B CN 107085850B CN 201710182739 A CN201710182739 A CN 201710182739A CN 107085850 B CN107085850 B CN 107085850B
- Authority
- CN
- China
- Prior art keywords
- image
- marked body
- foreign matter
- marked
- length
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 26
- 230000000873 masking effect Effects 0.000 title claims abstract description 9
- 238000003708 edge detection Methods 0.000 claims abstract description 11
- 239000000284 extract Substances 0.000 claims abstract description 5
- 238000012216 screening Methods 0.000 claims description 10
- 239000012141 concentrate Substances 0.000 claims description 2
- 238000007781 pre-processing Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims 1
- 238000010586 diagram Methods 0.000 description 13
- 239000004033 plastic Substances 0.000 description 6
- 239000004568 cement Substances 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000004364 calculation method Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 230000000717 retained effect Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 1
- 230000002146 bilateral effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 239000000571 coke Substances 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 239000012634 fragment Substances 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 230000001131 transforming effect Effects 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
- H04N1/46—Colour picture communication systems
- H04N1/56—Processing of colour picture signals
- H04N1/60—Colour correction or control
- H04N1/62—Retouching, i.e. modification of isolated colours only or in isolated picture areas only
- H04N1/626—Detection of non-electronic marks, e.g. fluorescent markers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/22—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition
- G06V10/225—Image preprocessing by selection of a specific region containing or referencing a pattern; Locating or processing of specific regions to guide the detection or recognition based on a marking or identifier characterising the area
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/107—Static hand or arm
- G06V40/11—Hand-related biometrics; Hand pose recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/13—Type of disclosure document
- G06V2201/131—Book
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of marked bodies of foreign matter in masking acquisition image, comprising: mark portion, surface have the two side's continuous patterns at least formed by one or more kinds of meta graph recognitions；Marked body is fixed on the foreign matter occurred in acquisition target by fixed part, so that foreign matter surface is covered by the mark portion in acquired image, is identified and is marked convenient for algorithm.A kind of method of foreign matter marked body in identification image includes the following steps: to carry out edge detection to flat image, obtains the edge graph in flat image；Extract whole profiles in the edge graph.A certain number of alternative straight sections are determined by algorithm, the region are determined according to the position of alternative every straightway, finally using the region of the similar area above or below marked body.A kind of book scanning method extends described to marked body image range, and the marked body image of current page is completed in removal eventually, completes the scanning of current page.
Description
Technical field
The present invention relates to a kind of marked bodies for the foreign matter that leaf through a book is covered in image collection type scanning process, identify in image
The algorithm and book scanning method of marked body, and the books page turning scan method of corresponding image collection type.It is related to patent point
Class-mark G06 is calculated；It calculates；G06F electricity Digital data processing G06F9/00 presetting apparatus is counted, for example, controller G06F9/
The program of 06 application deposit, i.e., it receives program using the storage inside of processing equipment and the G06F9/44 of program is kept to be used for
Execute the device of special procedure.
Background technique
Scanner based on video image acquisition passes through the camera being located above scanning object and acquires page photo, passes through view
Frequency is that scanning can be completed after algorithm handles image, eliminates and manually presses printed matter to be scanned in scanning surface
Traditional scanning mode bring hard work amount.
But such scanning means needs handspring in the thicker books of scanning thickness or during quick scanning
Page, and need the page with hand/finger pressing unilateral side or bilateral in order to keep page smooth during the scanning process, lead
The page image of algorithm acquisition is caused inevitably to have finger-image.
Area of skin color is extracted using oval complexion model to remove colour of skin near zone to position finger.The oval colour of skin
Model: transforming to YCrCb color space from rgb space for colour of skin image, and on the two-dimensional space of CrCb, sample areas presents ellipse
Round shape feature, thus people using a CrCb spatially approximate elliptic region as determine the colour of skin foundation.
But it is influenced due to the book contents multiplicity for scanning, and by light variation, is known with color characteristic merely
Other finger areas easily causes erroneous detection and missing inspection.
Summary of the invention
The it is proposed of the present invention in view of the above problems, and a kind of masking developed acquires the marked body of foreign matter in image, comprising:
Mark portion, surface have the two side's continuous patterns at least formed by one or more kinds of meta graph recognitions；Fixed part, will
Marked body is fixed on the foreign matter occurred in acquisition target, such as finger or the similar disclosed automatic page turning equipment appearance of flipbook
On the surface of the flipbook mechanism on page surface, so that foreign matter surface is covered by the mark portion in acquired image, it is convenient for
Algorithm identification and label.
For the ease of algorithm identification, as preferred embodiment, the pel includes the isometric straight line being parallel to each other
Section and quadrant or open circles.
For the ease of the pattern range of algorithm defined label body entirety, as preferred embodiment, two sides connect
Continuous pattern concentrates in the rectangle identification region that one is located in the middle part of mark portion, the length of each equal length segment and rectangular area
Side is vertical, and the elliptic focus (or central point of focal length) line is parallel with the long side of the rectangular area；
Further, in order to reinforce the contrast difference of pel and background, enhance edge variation gradient, to make pel
Can be more obvious under different light, as preferred embodiment, the color of the rectangle identification region is described two
The inverse of pel color in square continuous pattern.
Further, the calculation in order to make accurately is positioned and is identified (principle will be in algorithm portion details),
When the mark portion includes a variety of pels, each two side's continuous patterns are made of a kind of pel；Multiple two sides connect
Continuous pattern is parallel with rectangular area long side.
Including at least two side's continuous patterns of the Line Segment and four points in two side's continuous pattern of multirow
Two side's continuous patterns of circle/open circles.
The method of foreign matter marked body, includes the following steps: in a kind of identification image
Firstly, acquisition includes the flat image of the marked body；It completes to include at least binaryzation and the image of denoising is pre-
Processing.
Then, edge detection is carried out to flat image, obtains the edge graph in flat image；It extracts complete in the edge graph
Contouring.
By carrying out straight line screening to whole profiles, two side's connection figures of the Line Segment in the marked body are obtained
Case；
After again, the area image where Line Segment is extracted in the flat image of acquisition.
As preferred embodiment, a certain number of alternative straight sections are determined by algorithm first, according to every alternative
The position of straightway determines the region, is expanded outwardly using a biggish size by initial straight line fragment position, then
Each connected region (may be connection region caused by the overlapping of many straightway zone broadenings) after expansion is taken one
Method as a boundary rectangle obtains final area image.
By area image described in edge detection, the corresponding local edge figure of marked body is obtained, the local edge is extracted
It is alternately oval to obtain all elliptical local configurations by the screening to local configuration for local configuration in figure；It calculates every
The elliptic focus and length shaft length of a elliptic contour；
Examine the median of the neighbouring length and angle of straightway of each alternative circle (it is considered that detecting sometimes
In interference straightway on certain non-fingerstall patterns on side, such as the straight line and signage that occur in books compared with, maximum probability
The length and angle diversity ratio of straightway is larger, calculates average, it is most likely that sees in error, therefore uses median), lead to
It crosses by every straightway compared with angle and length median, removal deviation exceeds the alternative straight section of threshold range.
The position for considering Line Segment and elliptical pixel center respectively, selects the picture more close to the pattern of marked body
Two side's continuous patterns at plain center, calculate the pixel center of pattern, and in the embodiment of subsequent descriptions, Line Segment is more
Close to center, the pixel center C of Line Segment is selected, as marked body center benchmark；It calculates all elliptical flat
Equal long axis length R, the index as judge mark body apart from camera lens distance；The average length L for calculating all straight lines, as finger
Tilt down the foundation of size.
The image range of marked body is calculated according to the pixel center C, average major axis length R and average length L.
As preferred embodiment, edge detection is carried out to flat image and uses Canny edge detection；By determining to take turns
The length-width ratio of the external minimum rectangle size of wide surround the area threshold range, profile and boundary rectangle, rejects non-straight line segment profile.
In order to guarantee arithmetic accuracy, as preferred embodiment, final all qualified straightway pixels are calculated
Also there is alternative circle and alternative straight section corresponding relationship screening step before center.
Each straight line found is compared with each alternative circle, find each circumference enclose the condition of satisfaction association it is straight
Line.Usual marked body is remoter apart from camera lens, and the image pixel between point and point is apart from smaller；Marked body inclined degree is bigger, point with
Image pixel between point is apart from smaller.
Every straight line meets bound threshold requirement with respect to the proximal end of focus at a distance from focus, comprehensively considers acquisition page
Distance and marked body inclined degree and camera resolution of the height, marked body of image pickup head apart from camera lens.In conjunction with upper
Factor is stated, provides following parameter threshold: 8 pixel of lower limit, 45 pixel of the upper limit, meanwhile, distal end of the every straight line away from focus and coke
The distance of point meets threshold requirement: lower limit requirement: not less than 25 pixel of lower limit, can satisfy conventional mainstream resolution ratio 720p,
1080p and 2k even the Image Acquisition precision and the algorithm speed of service of 4k.
The association straight line that each circumference encloses is found according to the above method, is lower than if the circumference encloses the association straight line number found
4, then delete the alternative circle.
As preferred embodiment, determine before calculating final all qualified straightway pixel centers alternative straight
Whether line segment passes through corresponding alternative circle；If passing through, the alternative straight section is rejected.
Further, it is contemplated that in the books in practical situations, such as around the finger pressing of wear markers body
There may be certain circle (such as attached drawing) in appearance, the linear distance on round or Elliptical distance fingerstall is significantly greater than in marked body
Ellipse/circle, pretend as preferred embodiment, these external interferences circle weeded out using threshold value: calculating final all symbols
Before the straightway pixel center of conjunction condition, the line of alternative circle row where whether the alternative circular focus deviates is determined；If partially
From being more than threshold distance, then the alternative circle is rejected.
Further, calculate before calculating final all qualified straightway pixel centers relevant straight line
Central line degree, i.e. every straight central to relevant straight line the center line being centrally formed average distance；
If above-mentioned average distance is greater than 3 pixels, it is associated with straight line group and is deleted, while corresponding alternative circle is deleted.
A kind of book scanning method includes the following steps:-is directed to the book pages of the marked body with marked body masking
Two dimensional image, determine the image range of marked body；- using the region of the similar area above or below marked body, it extends to
The image range of the marked body, removal complete the marked body image of current page, complete the scanning of current page.
Detailed description of the invention
For the clearer technical solution for illustrating the embodiment of the present invention or the prior art, to embodiment or will show below
There is attached drawing needed in technical description to do one simply to introduce, it should be apparent that, the accompanying drawings in the following description is only
Some embodiments of the present invention without creative efforts, may be used also for those of ordinary skill in the art
To obtain other drawings based on these drawings.
Fig. 1 is the fingerstall schematic diagram in the embodiment of the present invention as marked body
Fig. 2 is straightway pattern schematic diagram in the embodiment of the present invention
Fig. 3 is quadrant pattern schematic diagram in the embodiment of the present invention
Fig. 4 is preferred pattern schematic diagram in the embodiment of the present invention
Fig. 5 is flat image schematic diagram of the present invention
Fig. 6 is topography's schematic diagram that the present invention extracts
Fig. 7 is marked body (fingerstall) image range exposure mask figure in inventive algorithm embodiment
Fig. 8 is exposure mask analysis diagram in inventive algorithm embodiment
Fig. 9 is the detail structure chart of exposure mask in inventive algorithm embodiment
Figure 10 is that marked body parameter calculates schematic diagram in inventive algorithm embodiment
Figure 11 is the scan image schematic diagram that algorithm eliminates after marked body in inventive algorithm embodiment
Figure 12 is the application scenarios schematic diagram of inventive algorithm embodiment 2
Figure 13 is recognition result schematic diagram in inventive algorithm embodiment 2
Figure 14 is image-forming principle schematic diagram in inventive algorithm embodiment
Figure 15 is the algorithm flow chart of book scanning of the present invention
Specific embodiment
To keep the purposes, technical schemes and advantages of the embodiment of the present invention clearer, below with reference to the embodiment of the present invention
In attached drawing, technical solution in the embodiment of the present invention carries out clear and complete description:
As shown in Figure 1, what is provided in the present embodiment is for the fingerstall scheme using finger flipbook, the fixed part is
Circle is similar to plastic cement/rubber fingerstall of revolving body, and convenient for being socketed on user's flipbook finger, while plastic cement has with papery page
Biggish frictional force.
At the middle part of figure as shown in Figure 1, equipped with the rectangle identification region perpendicular to central axes, itself is equipped in region
Two enough side's continuous patterns of the isometric straightway being parallel to each other, and the two row two side continuous pattern enough by open circles, two
Row open circles are staggered.The scheme of the circle of elliptical special case is only considered as in the present embodiment.
In color selection, (i.e. parallel equal length segment and open circles is white, the background of rectangle identification region to pel
Color is inverse-black of white), fingerstall generally yellow (as preferred embodiment, can also be used and books papery area
Not biggish other colors).
In order to facilitate wearing, be additionally provided with the rubber/plastic tooth of dense arrangement in fingerstall/marked body inner surface, with rubber/
The elastic collective effect of plastic cement material guarantees the secured of wearing, while will not generate excessive pressure to finger and cause discomfort
(when finger diameter dimension is larger, rubber teeth itself can deformation occurs, reduces to finger pressure).
Meanwhile rubber/plastic tooth enters fingerstall direction along finger and is arranged, and convenient for wearing and taking off fingerstall, while guaranteeing hand
Refer to that the fastness during transverse movement flipbook, the especially frictional force after finger perspiration between rubber material are likely less than
The frictional force of papery page, the case where causing fingerstall to deflect.
Embodiment 1, book scanning application scenarios, the present embodiment mainly solve in image collection type scanning process, finger etc.
Influence during page turning to books image, as shown in Fig. 2-15:
(1) straightway of suitable size is found from image:
1. image is switched to grayscale image；
2. carrying out the median filtering denoising of [5,5] size；
3. carrying out Canny edge detection, gradient lower limit 75, the upper limit 120 obtains the edge graph of image；
4. extracting profile from edge graph；
5. analyzing whether each little profile meets the size and shape requirement of fingerstall straight line, to reject non-straight line segment
Profile:
A) area that profile surrounds need to meet bound requirement: 10 pixel of area lower limit, 500 pixel of the area upper limit；
B) ask the minimum circumscribed rectangle of profile corresponding wide and high, it is desirable that the long side of boundary rectangle needs to meet straight length
Bound requirement: length lower limit: 12 pixels, length limit: 70 pixels；
C) length-width ratio of boundary rectangle need to be greater than 3；
For the profile for meeting above-mentioned geomery requirement, two ends of corresponding straightway are acquired according to its boundary rectangle
Point, by each straightway indicated with two endpoints, alternately straightway is stored.
6. the rejecting that pair alternative straight found carries out double line: if judgment basis is two straight line two endpoints
Distance both less than 3 pixels, then it is assumed that they are overlapped.
Yellow line segment in this figure is the qualified straight line that the step is found.
(2) fingerstall regional area that may be present is positioned according to straightway closeness
1. traversing every straightway, the center of the straightway, slope and length is calculated；
2. the straightway is successively compared with the feature of other every straightway, both see whether to meet following all want
It asks:
A) the distance between the center of two straightways meets bound requirement: 4 pixels of lower limit, 60 pixel of the upper limit；
B) the slope difference of two straightways is less than 0.05；
C) difference in length of two straightways is less than 0.3 times of any length of straigh line.
If the relationship of certain straightway and the straightway meets all above requirement, the straightway is corresponding similar straight
Line segment, which counts, increases by 1.
3. the region for being more than threshold value 5 for similar straight line accumulative total extracts, general fingerstall region is obtained.
(3) circle is found in above-mentioned zone
1. local positioning figure above is transformed to grayscale image；
2. pair grayscale image carries out Canny edge detection, gradient lower limit 50, the upper limit 100 obtains edge graph；
3. extracting profile for edge graph；
4. pair each profile is analyzed, to search the profile for meeting oval feature:
A) area of profile need to meet threshold range requirement: 200 pixel of lower limit, 2500 pixel of the upper limit；
B) according to the two-dimentional point set fitted ellipse of profile, the area of fitted ellipse and the difference of actual profile area, the two are asked
Difference in areas needs to meet: less than 10 pixels；Otherwise it is assumed that the ovality of the profile is insufficient.
The profile for meeting both the above condition is retained, it is believed that they are alternative fingerstall circular patterns, and are recorded every
Elliptic focus, the length shaft length of a profile.
The part of figure blue is the alternative circle found.
(4) circle is combined with straight line information, alternative circle and straight line is screened
1. screening for the first time:
A) each straight line found is compared with each alternative circle, find each circumference enclose the condition of satisfaction association it is straight
Line.Condition is every straight line meets bound requirement: 8 pixel of lower limit, 45 picture of the upper limit with respect to the proximal end of focus at a distance from focus
Element；Meanwhile distal end of the every straight line away from focus meets lower limit requirement at a distance from focus: being not less than 25 pixel of lower limit；
B) the association straight line that each circumference encloses is found according to the above method, if to enclose the association straight line number found low for the circumference
In 4, then the alternative circle is deleted；
C) by the remaining angle, length and central store for alternatively justifying surrounding associated straight line for meeting condition.
2. programmed screening:
A) median for calculating the neighbouring length and angle of straight line of each alternative circle, by the angle of every straight line and length with
The two medians compare, and difference need to meet bound requirement: the angle and median angle degree difference of every straight line cannot be greater than 5
Degree；The difference of every line and middle bit length cannot be greater than the 5% of middle bit length；
B) straight line for being unsatisfactory for bound requirement is deleted from association rectilinear alignments.
If c) screened by upper step, some circumference encloses remaining association straight line number if it is less than 4, then it is standby to delete this
Choosing circle.
3. third time is screened:
A) otherwise two endpoints of every straight line illustrate that this is passed in straight line through not across the both ends associated with it alternatively justified
The circle, and have the circle passed in straight line through that will be deleted.
4. the 4th screening:
A) calculate alternatively circle relevant straight line central point, fit a center line using these central points；
B) each center alternatively justified is calculated at a distance from above-mentioned center line, which cannot be less than in circle association straight line
0.8 times of bit length.The circle that condition is not satisfied is deleted.
5. the 5th screening:
A) calculate relevant straight line central line degree, i.e. every straight central to relevant straight line be centrally formed
Center line average distance；
If b) above-mentioned average distance is greater than 3 pixels, it is associated with straight line group and is deleted, while corresponding alternative circle is deleted.
Through above-mentioned totally five wheel screening, the circle on fingerstall that retained circle as finally determines, while also saving each
The information of the corresponding association straight line for meeting condition of circle.
After completing positioning, you can get it complete fingerstall image masks, as shown in Figure 10, typical exposure mask range is by such as
Lower four parts composition: a rectangle (for limiting fingerstall width), the ellipse of finger middle section, finger fingertip part it is ellipse
Round and Fingers root portion ellipse.
For three elliptical short axles all along finger A-B line direction, the direction corresponds to the length or length of finger areas in figure
Spend direction.The length of three ellipse short shafts all relies on L: when finger tilt degree is certain, and it is closer apart from camera lens when, L becomes larger, this
When finger areas length increase；When finger is certain with distance of camera lens, finger fingertip tilts down that amplitude is bigger, and L is smaller, at this time hand
Refer to that zone length is shorter.
Diagram three elliptical long axis and rectangle width be all perpendicular to A-B line (i.e. the central axial direction of fingerstall,
In the non-use state, cross section is integrally in symmetrical image to rubber/plastic fingerstall, similar oval) direction, this direction corresponds to hand
Refer to the width in region.The length of the amount all relies on R, because R is because be elliptical long axis, with finger tilt size
Unrelated, the distance only with finger apart from camera lens is related.When finger is closer apart from camera lens, R becomes larger, and finger width increases at this time.
Wherein the size of each part and position correspond to fingerstall center C, ellipse average major axis length R and straight line average length
The relationship of L is as follows:
Rectangle: centered on C, long 15, it is wide by 9.5；The ellipse of finger middle section: centered on C, minor axis radius 4.3 is long
Axis radius 7；The ellipse of tip portion: centered on A, | A-C |=3L, minor axis radius 3, major axis radius 3.5；Refer to the ellipse of root portion
Circle: centered on B, | B-C |=5L, minor axis radius 5, major axis radius 6.
It is as shown in figure 15: for mathematics relationship of the present invention, in general, the line of equal length in the picture, corresponding
Practical parallel lines length camera photocentre corresponding thereto distance it is directly proportional；And the parallel lines of upper equal length in kind,
The vertical range of corresponding length optical center corresponding thereto is inversely proportional on image.
Embodiment 2, differentiation of the marked body as the application scenarios generally identified, for bulk article.In mouse upper surface
It is provided with marked body.In the present embodiment, the fixed part of marked body is preferably sticker form.By using the calculation in embodiment 1
Method can recognize marked body.
Due to the primitive features and inverse feature of marked body itself, so that under the application scenarios of multiple color background,
The identification of all achievable marked body guarantees accuracy of identification.
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto,
Anyone skilled in the art in the technical scope disclosed by the present invention, according to the technique and scheme of the present invention and its
Inventive concept is subject to equivalent substitution or change, should be covered by the protection scope of the present invention.
Claims (9)
1. the marked body of foreign matter in a kind of masking acquisition image, characterized by comprising:
Mark portion, surface have the two side's continuous patterns at least formed by one or more kinds of meta graph recognitions；
Marked body is fixed on the surface of the foreign matter occurred in acquisition target, so that foreign matter table in acquired image by fixed part
Face is covered by the mark portion；
When the mark portion includes a variety of pels, each two side's continuous patterns are made of a kind of pel；Described
Two side's continuous patterns concentrate in the rectangle identification region that one is located in the middle part of mark portion, multiple two sides continuous patterns and rectangular area
Long side is parallel.
2. the marked body of foreign matter in masking acquisition image according to claim 1, it is further characterized in that the pel packet
Include the isometric straightway and ellipse being parallel to each other.
3. the marked body of foreign matter in masking acquisition image according to claim 1 or 2, it is further characterized in that each isometric straight
Line segment is vertical with the long side of rectangular area, the long side of the midpoint line and the rectangular area of elliptic focus or oval focal length
In parallel；The color of the rectangle identification region is the inverse of pel color in the two sides continuous pattern.
4. the marked body of foreign matter in masking acquisition image according to claim 1, it is further characterized in that two side of multirow is continuous
Two side's continuous pattern of two side's continuous patterns and quadrant/open circles of Line Segment is included at least in pattern.
5. a kind of method of foreign matter marked body in identification image, it is characterised in that include the following steps:
- acquisition includes the flat two-dimensional images of marked body as claimed in claim 4；It completes to include at least binaryzation and denoising
Image preprocessing；
- edge detection is carried out to flat image, obtain the edge graph in flat image；Extract whole profiles in the edge graph；
- by carrying out straight line screening to whole profiles, obtain alternative parallel straightway；
- the area image in the flat image of acquisition where extraction Line Segment；
- by area image described in binaryzation and edge detection, local edge figure is obtained, is extracted in the local edge figure
Local configuration is obtained all elliptical local configurations and is alternately justified by the screening to local configuration；Calculate each elliptic wheel
Wide elliptic focus and length shaft length；
- medians of the neighbouring length and angle of straightway of each alternative circle is examined, by by every straightway and angle
Compare with length median, removal deviation exceeds the alternative straight section of threshold range；
The pixel center C of the final all qualified straightways of-calculating, as marked body center benchmark；It calculates all
Elliptical average major axis length R, the index as judge mark body apart from camera lens distance；The average length L of all straight lines is calculated,
The foundation of size is tilted down as finger；
- image range of marked body is calculated according to the pixel center C, average major axis length R and average length L.
6. it is according to claim 5 identification image in foreign matter marked body method, it is further characterized in that flat image into
Row edge detection uses Canny edge detection；By determining profile surround the area threshold range, the external minimum rectangle size of profile
And the length-width ratio of boundary rectangle, reject non-straight line segment profile.
7. the method for foreign matter marked body in identification image according to claim 5, it is further characterized in that calculating final all
Determine whether alternative straight section passes through corresponding alternative circle before qualified straightway pixel center；If passing through, reject
The alternative straight section；
Before calculating final all qualified straightway pixel centers, determine standby where whether the alternative circular focus deviates
Select the line of circle row；If deviateing is more than threshold distance, the alternative circle is rejected.
8. the method for foreign matter marked body in identification image according to claim 5, it is further characterized in that calculating final all
Calculate the central line degree of the relevant straight line of institute before qualified straightway pixel center, i.e. every straight central is to owning
It is associated with the average distance for the center line of straight line being centrally formed；
If above-mentioned average distance is greater than 3 pixels, it is associated with straight line group and is deleted, while corresponding alternative circle is deleted.
9. a kind of book scanning method, it is characterised in that include the following steps:
- for the two dimensional image for the book pages for having the marked body that marked body covers, it is any one using such as claim 5-8
Method described in item claim, determines the image range of marked body；
- using the region of the similar area above or below marked body, extend to the image range of the marked body, removal
The marked body image for completing current page, completes the scanning of current page.
Priority Applications (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710182739.2A CN107085850B (en) | 2017-03-24 | 2017-03-24 | The method and book scanning method that masking acquires the marked body of foreign matter in image, identifies foreign matter marked body in image |
| PCT/CN2017/078745 WO2018170937A1 (en) | 2017-03-24 | 2017-03-30 | Marker for occluding foreign matter in acquired image, method for recognizing foreign matter marker in image and book scanning method |
| US16/496,877 US10846549B2 (en) | 2017-03-24 | 2017-03-30 | Marker for occluding foreign matter in acquired image, method for recognizing foreign matter marker in image and book scanning method |
| EP17901740.5A EP3605461A4 (en) | 2017-03-24 | 2017-03-30 | Marker for occluding foreign matter in acquired image, method for recognizing foreign matter marker in image and book scanning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710182739.2A CN107085850B (en) | 2017-03-24 | 2017-03-24 | The method and book scanning method that masking acquires the marked body of foreign matter in image, identifies foreign matter marked body in image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107085850A CN107085850A (en) | 2017-08-22 |
| CN107085850B true CN107085850B (en) | 2019-10-01 |
Family
ID=59614221
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710182739.2A Active CN107085850B (en) | 2017-03-24 | 2017-03-24 | The method and book scanning method that masking acquires the marked body of foreign matter in image, identifies foreign matter marked body in image |
Country Status (4)
| Country | Link |
|---|---|
| US (1) | US10846549B2 (en) |
| EP (1) | EP3605461A4 (en) |
| CN (1) | CN107085850B (en) |
| WO (1) | WO2018170937A1 (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110910393B (en) * | 2018-09-18 | 2023-03-24 | 北京市商汤科技开发有限公司 | Data processing method and device, electronic equipment and storage medium |
| CN109712132B (en) * | 2018-12-28 | 2021-03-19 | 中国科学院长春光学精密机械与物理研究所 | Method for automatically screening and counting by using electronic device counting device |
| US20200320334A1 (en) * | 2019-04-05 | 2020-10-08 | Ent. Services Development Corporation Lp | Systems and methods for digital image-based object authentication |
| CN111079494B (en) * | 2019-06-09 | 2023-08-25 | 广东小天才科技有限公司 | Learning content pushing method and electronic equipment |
| CN111814686A (en) * | 2020-07-09 | 2020-10-23 | 国网山西省电力公司吕梁供电公司 | Vision-based power transmission line identification and foreign matter invasion online detection method |
| US12046046B2 (en) * | 2021-04-09 | 2024-07-23 | Magna Electronics, Llc | Tagging objects with fault override patterns during calibration of vehicle sensing systems |
| CN113361643A (en) * | 2021-07-02 | 2021-09-07 | 人民中科(济南)智能技术有限公司 | Deep learning-based universal mark identification method, system, equipment and storage medium |
| CN114612496B (en) * | 2022-03-18 | 2024-05-31 | 易思维(杭州)科技股份有限公司 | Straight line segment contour extraction method |
| CN115147407B (en) * | 2022-08-29 | 2022-11-18 | 聊城市博源节能科技有限公司 | Bearing quality detection method based on computer vision |
| CN117066702B (en) * | 2023-08-25 | 2024-04-19 | 上海频准激光科技有限公司 | Laser marking control system based on laser |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1753448A (en) * | 2004-09-21 | 2006-03-29 | 兄弟工业株式会社 | Image processing apparatus |
| CN101295149A (en) * | 2007-04-24 | 2008-10-29 | 京瓷美达株式会社 | Image forming apparatus |
| TW201023622A (en) * | 2008-12-09 | 2010-06-16 | Qisda Corp | Scanning method and auxiliary apparatus of auto-executing appending scanning function |
| CN102376566A (en) * | 2010-08-19 | 2012-03-14 | 三星电子株式会社 | Method of forming a pattern structure for a semiconductor device |
| CN102646193A (en) * | 2012-03-23 | 2012-08-22 | 河海大学 | Segmentation method of character images distributed in ring shape |
| CN103377462A (en) * | 2012-04-16 | 2013-10-30 | 富士通株式会社 | Method and device for processing scanned image |
| CN104509102A (en) * | 2012-07-27 | 2015-04-08 | 日产自动车株式会社 | Three-dimensional object detection device and foreign object detection device |
| CN106199750A (en) * | 2015-05-04 | 2016-12-07 | 中国科学院上海高等研究院 | Flying Area in Airport activity foreign matter detection system and detection method thereof |
Family Cites Families (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5065257A (en) * | 1985-03-20 | 1991-11-12 | Canon Kabushiki Kaisha | Image processing apparatus with means for compensating for dust., etc., on an original |
| JP3436025B2 (en) * | 1995-12-27 | 2003-08-11 | ミノルタ株式会社 | Correction method of read image and image reading device |
| US20090254994A1 (en) * | 2002-02-18 | 2009-10-08 | David Lynch Waterson | Security methods and systems |
| JP2004357190A (en) * | 2003-05-30 | 2004-12-16 | Ricoh Co Ltd | Image reading apparatus |
| US10635723B2 (en) * | 2004-02-15 | 2020-04-28 | Google Llc | Search engines and systems with handheld document data capture devices |
| JP2008227889A (en) * | 2007-03-13 | 2008-09-25 | Ricoh Co Ltd | Image reader and control method therefor |
| US8018631B2 (en) * | 2007-07-04 | 2011-09-13 | Ricoh Company, Limited | Original reading device and image forming apparatus |
| JP2010283789A (en) * | 2009-06-08 | 2010-12-16 | Canon Inc | Image reading apparatus |
| CN102034445B (en) * | 2009-09-28 | 2013-01-09 | 群康科技(深圳)有限公司 | Power supply circuit of liquid crystal display |
| CN101915573B (en) | 2010-08-04 | 2012-12-12 | 中国科学院自动化研究所 | Positioning measurement method based on key point detection of marker |
| CA2849560A1 (en) * | 2013-04-22 | 2014-10-22 | Pierre Hamel | Optical character recognition of text in an image according to a prioritized processing sequence |
| JP6099686B2 (en) * | 2015-03-20 | 2017-03-22 | 株式会社Pfu | Image processing apparatus, region detection method, and computer program |
| CN105872291B (en) * | 2016-05-31 | 2018-11-27 | 大连成者科技有限公司 | Intelligent Web high definition scanner with laser-adjusting |
| CN106127203B (en) | 2016-06-29 | 2019-06-25 | 孟祥雨 | It is a kind of that knowledge method for distinguishing being carried out to image using object positioning and follow-up mechanism |
| CN106412570B (en) | 2016-10-11 | 2018-08-07 | 惠州市德赛西威汽车电子股份有限公司 | A kind of assay method and tools for measurement of camera system noise removal capability |
| CN107291993A (en) * | 2017-05-27 | 2017-10-24 | 中国石油大学(华东) | The analogy method of pre-crosslinked gel suspension micro flow in a kind of porous media |
| US10291805B1 (en) * | 2018-03-20 | 2019-05-14 | Kabushiki Kaisha Toshiba | Image processing apparatus |
| WO2020076472A2 (en) * | 2018-09-19 | 2020-04-16 | Sana Rezgui | Human-robots: the new specie |
-
2017
- 2017-03-24 CN CN201710182739.2A patent/CN107085850B/en active Active
- 2017-03-30 WO PCT/CN2017/078745 patent/WO2018170937A1/en active Application Filing
- 2017-03-30 US US16/496,877 patent/US10846549B2/en active Active
- 2017-03-30 EP EP17901740.5A patent/EP3605461A4/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1753448A (en) * | 2004-09-21 | 2006-03-29 | 兄弟工业株式会社 | Image processing apparatus |
| CN101295149A (en) * | 2007-04-24 | 2008-10-29 | 京瓷美达株式会社 | Image forming apparatus |
| TW201023622A (en) * | 2008-12-09 | 2010-06-16 | Qisda Corp | Scanning method and auxiliary apparatus of auto-executing appending scanning function |
| CN102376566A (en) * | 2010-08-19 | 2012-03-14 | 三星电子株式会社 | Method of forming a pattern structure for a semiconductor device |
| CN102646193A (en) * | 2012-03-23 | 2012-08-22 | 河海大学 | Segmentation method of character images distributed in ring shape |
| CN103377462A (en) * | 2012-04-16 | 2013-10-30 | 富士通株式会社 | Method and device for processing scanned image |
| CN104509102A (en) * | 2012-07-27 | 2015-04-08 | 日产自动车株式会社 | Three-dimensional object detection device and foreign object detection device |
| CN106199750A (en) * | 2015-05-04 | 2016-12-07 | 中国科学院上海高等研究院 | Flying Area in Airport activity foreign matter detection system and detection method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| US20200104621A1 (en) | 2020-04-02 |
| EP3605461A1 (en) | 2020-02-05 |
| WO2018170937A1 (en) | 2018-09-27 |
| US10846549B2 (en) | 2020-11-24 |
| CN107085850A (en) | 2017-08-22 |
| EP3605461A4 (en) | 2021-03-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107085850B (en) | The method and book scanning method that masking acquires the marked body of foreign matter in image, identifies foreign matter marked body in image | |
| CN112508826A (en) | Printed matter defect detection method based on feature registration and gradient shape matching fusion | |
| US8073233B2 (en) | Image processor, microscope system, and area specifying program | |
| TWI467515B (en) | Multi-color dropout for scanned document | |
| JP5351673B2 (en) | Appearance inspection device, appearance inspection method | |
| US8019132B2 (en) | System for recognizing fingerprint image, method and program for the same | |
| JPH0877334A (en) | Automatic feature point extracting method for face image | |
| US20150146943A1 (en) | Method of recognizing contactless fingerprint and electronic device for performing the same | |
| JP3830998B2 (en) | Ruled line removal method and character recognition apparatus using the same | |
| CN114863492B (en) | Method and device for repairing low-quality fingerprint image | |
| CN107392849A (en) | Target identification and localization method based on image subdivision | |
| JP5769559B2 (en) | Image processing apparatus, image processing program, robot apparatus, and image processing method | |
| AU2021368390B2 (en) | Multi-target recognition system and method for follow-up robot based on coded thermal infrared mark | |
| US20050271260A1 (en) | Device, method and program for removing pores | |
| CN108701204A (en) | A kind of method and device of one-dimension code positioning | |
| Ruslau et al. | Edge detection of digital image with different edge types | |
| CN109523594A (en) | A kind of vision tray characteristic point coordinate location method and system | |
| JP4020202B2 (en) | Striped pattern extraction system and striped pattern extraction method | |
| JP4991278B2 (en) | Image processing device | |
| JP2006163662A (en) | Device and method for recognizing number of fingers | |
| KR102178636B1 (en) | Finger segmentation method | |
| CN109376730B (en) | Gesture recognition method and device | |
| JPH08161508A (en) | Pattern detecting method | |
| CN109543531A (en) | A kind of method of fiber extraction and fiber vision correcting | |
| Qian et al. | Separation of high-resolution samples of overlapping latent fingerprints using relaxation labeling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| PE01 | Entry into force of the registration of the contract for pledge of patent right |
Denomination of invention: Masking the foreign object marker in the collected image, the method of recognizing the foreign object marker in the image, and the book scanning method Effective date of registration: 20220930 Granted publication date: 20191001 Pledgee: Industrial Bank Limited by Share Ltd. Dalian branch Pledgor: DALIAN CZUR TECH Co.,Ltd. Registration number: Y2022980017426 |
|
| PE01 | Entry into force of the registration of the contract for pledge of patent right | ||
| PC01 | Cancellation of the registration of the contract for pledge of patent right |
Date of cancellation: 20231018 Granted publication date: 20191001 Pledgee: Industrial Bank Limited by Share Ltd. Dalian branch Pledgor: DALIAN CZUR TECH Co.,Ltd. Registration number: Y2022980017426 |
|
| PC01 | Cancellation of the registration of the contract for pledge of patent right |